Population Comparison of Right Whale Body Condition Reveals Poor State of the North Atlantic Right Whale

Vol. 640: 1–16, 2020 MARINE ECOLOGY PROGRESS SERIES Published April 23 https://doi.org/10.3354/meps13299 Mar Ecol Prog Ser

OPENPEN ACCESSCCESS FEATURE ARTICLE Population comparison of right whale body condition reveals poor state of the North Atlantic right whale

Fredrik Christiansen1,2,3,*, Stephen M. Dawson, John W. Durban, Holly Fearnbach, Carolyn A. Miller, Lars Bejder, Marcela Uhart, Mariano Sironi, Peter Corkeron, William Rayment, Eva Leunissen, Eashani Haria, Rhianne Ward, Hunter A. Warick, Iain Kerr, Morgan S. Lynn, Heather M. Pettis, Michael J. Moore

ABSTRACT: The North Atlantic right whale Eubala - ena glacialis (NARW), currently numbering <410 individuals, is on a trajectory to extinction. Al though direct mortality from ship strikes and fishing gear entanglements remain the major threats to the population, reproductive failure, resulting from poor body condition and sublethal chronic entanglement stress, is believed to play a crucial role in the population decline. Using photo grammetry from unmanned aerial vehicles, we conducted the largest population assessment of right whale body condition to date, to determine if the condition of NARWs was poorer than 3 seemingly healthy (i.e. growing) populations of southern right whales E. australis (SRWs) in Argen - Three healthy southern right whales (left three photographs) tina, Australia and New Zealand. We found that next to a North Atlantic right whale (right) in visibly poorer body condition NARW juveniles, adults and lactating females all had lower body condition scores compared to the SRW Photos: Fredrik Christiansen (left & center-left), Stephen M. Dawson (center-right), John W. Durban/Holly Fearnbach (right) populations. While some of the difference could be the result of genetic isolation and adaptations to local environmental conditions, the magnitude suggests KEY WORDS: Baleen whale · Bioenergetics · that NARWs are in poor condition, which could be Eubalaena · Morphometrics · Photogrammetry · suppressing their growth, survival, age of sexual Unmanned aerial vehicles maturation and calving rates. NARW calves were found to be in good condition. Their body length, however, was strongly determined by the body condition of their mothers, suggesting that the poor condition of lactating NARW females may cause a 1. INTRODUCTION reduction in calf growth rates. This could potentially lead to a reduction in calf survival or an in crease in Right whale populations around the world were se- female calving intervals. Hence, the poor body con- verely depleted (some reduced to less than 5% of the dition of individuals within the NARW population is original population size) by commercial whaling op- of major concern for its future viability. erations from the 11th to the 20th century in the North Atlantic (Aguilar 1986, Reeves et al. 1999) and during

© The authors 2020. Open Access under Creative Commons by *Corresponding author: [email protected] Attribution Licence. Use, distribution and reproduction are un - restricted. Authors and original publication must be credited. Publisher: Inter-Research · www.int-res.com 2 Mar Ecol Prog Ser 640: 1–16, 2020

the 19th to 20th century in the Southern hemisphere (Christiansen et al. 2016a). Like most baleen whales, (Dawbin 1986, Jackson et al. 2016), as well as by ille- right whales make annual migrations between high- gal Soviet whaling in the Southern hemisphere in the latitude feeding grounds in summer and low-latitude 1950s to early 1970s (Yablokov 1994, Tormosov et al. breeding grounds in winter (Bannister et al. 1999). 1998). Since the cessation of commercial whaling, the Females become sexually mature at around 9 yr old southern right whale Eubalaena australis (SRW) has and give birth to a single calf at a time (Kraus et al. been recovering at a relatively rapid pace throughout 2001, Cooke et al. 2003, Burnell 2008). They are ‘cap- most of its range, and currently numbers in the tens ital’ breeders, fasting during the winter breeding of thousands globally (IWC 2013). The population season, and thus have a finite amount of energy to growth rate during this time has been as high as 5.55% invest in late pregnancy and lactation (Lockyer 1987, for Australia (Bannister 2016), 5−7% for New Zealand Stephens et al. 2009). Christiansen et al. (2018) (Carroll et al. 2013) and 6.5% for Argentina (Cooke et showed in SRWs that maternal size (body length and al. 2015), although the growth rate of the latter has condition) has a direct effect on the amount of energy slowed down substantially (Crespo et al. 2019). that lactating females invest in their calves, which in In contrast, the recovery of the North Atlantic right turn dictates calf growth rates. When conditions are whale E. glacialis (NARW) has been considerably favourable, females generally have a 3 yr reproduc- slower, with a mean annual growth rate of 2.8% be- tive cycle consisting of 1 yr of gestation, 1 yr of lacta- tween 1990 and 2010 (Pace et al. 2017). More recent tion and 1 yr of resting (to recover energy stores) abundance estimates, between 2010 and 2015, indicate (Best 1994). The mean calving interval for SRWs is that the population has been declining at a rate of just close to this 3 yr minimum, at 3.33 yr in Australia under 1% per year (Pace et al. 2017). The rate of (Burnell 2001), 3.31 yr in New Zealand (Davidson et decline has been higher for females, which dropped at al. 2017) and between 2.96 and 3.24 yr in Argentina approximately 7% between 2010 and 2015, compared (Marón et al. 2015). In contrast, since 2015, the mean to about 4% for males over the same period (Pace et al. calving interval for NARW females is >7 yr (Pettis et 2017). The situation for NARWs was further worsened al. 2020), suggesting that they need several years by an unusual mortality event between 1 November longer to recover from a reproductive event. Apart 2016 and 31 December 2017, when at least 17 juvenile from body condition having a direct effect on female and adult right whales died as a result of entangle- reproductive success, it can also influence juvenile ments and vessel strikes (NARWC 2018). In December growth rates (Douhard et al. 2017) and the age of 2015, prior to the 2017 mortalities, the species’ abun- sexual maturation (Sigurjónsson et al. 1990), which dance was estimated at 451 individuals, of which 186 could negatively influence population growth. were females. The best estimate as of the end of 2017 The aim of this study was to assess the body condi- was 411 animals (Pettis et al. 2018). While fishing gear tion of the NARW. Although the population’s body entanglements and ship strikes are the largest direct condition (based on visual assessment) has declined anthropogenic threats to the NARW population during the last 3 decades (Rolland et al. 2016), a com- (Moore et al. 2004, Knowlton et al. 2012, van der Hoop parison to healthy (growing) right whale populations et al. 2013, Kraus et al. 2016), reduced reproductive is needed to assess its current status. Unfortunately, rate resulting from nutritional stress (i.e. poor body no historical data on NARW body condition exist to condition) has been hypothesised as a factor further allow such a comparison. Instead, the best opportu- contributing to the population decline (Kraus et al. nity to assess the relative body condition of NARW 2001, Reeves et al. 2001, Schick et al. 2013, Rolland et comes from a comparison with their closest living rel- al. 2016). The sublethal impacts of entanglement have ative, the SRW. We therefore compared the body con- also been modelled to significantly impact reproduc- dition of NARWs with 3 seemingly healthy (i.e. grow- tive success (van der Hoop et al. 2017). ing) populations of SRW in Australia, New Zealand The effect of body condition on reproduction is well and Argentina. Although we were comparing 2 dif- documented in both terrestrial (Albon et al. 1983, ferent species of right whales (Rosenbaum et al. 2000, Loudon et al. 1983, Skogland 1984, Atkinson & Gaines et al. 2005), which might differ in their body Ramsay 1995, Festa-Bianchet 1998) and marine condition due to genetic differences, our rationale mammals (Arnbom et al. 1997, Boltnev & York 2001, was that body condition, similar to most traits closely Bowen et al. 2001, Wheatley et al. 2006). In baleen associated with fitness, shows low genetic variance whales, female body condition influences fecundity relative to environmental variance (Mousseau & Roff (Lockyer 2007, Williams et al. 2013), foetal growth 1987, Kruuk et al. 2000). We also show that NARWs (Christiansen et al. 2014) and calf body condition and SRWs are very similar in body shape, size and life Christiansen et al.: Body condition of right whale populations 3

history characteristics, which should facilitate com- North Atlantic, Australia, New Zealand and Argen - parison. Based on the lower population growth rate tina (Fig. 1). Photographs of NARW lactating females and longer calving interval of the NARW, our main and calves were collected on their calving grounds in hypothesis is that NARWs are in poorer body condi- Florida, USA, between 12 January and 22 February tion compared to SRWs. To help infer the potential ef- 2016 and 2017, while juvenile and adult NARWs fects of reduced body condition on different life his- were photographed on their feeding grounds in Cape tory parameters, we split our analysis into different Cod Bay, USA, between 21 March and 27 April 2016 reproductive classes (calves, juveniles, adults and and 2017 (Fig. 1). All SRWs were measured on their lactating females). We expected lactating females to breeding grounds at the Head of Bight, Australia, overall have a higher body condition relative to the between 25 June and 25 September 2016, the Auck- other reproductive classes, since they must have land Islands, New Zealand, between 28 July and 14 had sufficient energy reserves to complete gestation August 2016, and in Península Valdés, Argentina, (Lockyer 1981, Christiansen et al. 2014). However, between 3 August and 12 November 2018. with NARWs being affected by numerous anthro- In each location, multirotor UAVs were flown pogenic factors, we expected the body condition of from either land (Australia and Argentina) and/or lactating females to be significantly lower than for boats (North Atlantic, New Zealand and Argentina) SRWs. In baleen whales, a lower maternal body con- above a surfacing whale at altitudes between 17.8 dition has been shown to negatively influence calf and 55.1 m (mean = 31.3 m, SD = 8.01; Argentina = growth rates (Christiansen et al. 2018) and body con- 17.8−37.0 m; Australia = 27.9−46.6 m; New Zealand = dition (Christiansen et al. 2016a). We therefore antici- 17.9−51.3 m; North Atlantic = 26.8−55.1 m), and ver- pated NARW calves to have a poorer body condition tical photographs were taken of the dorsal side of the and/or a smaller body size (i.e. length) compared to whale (Fig. 2A). For the North Atlantic study site, an SRW calves. Finally, with fishing gear entanglements APH-22 hexacopter with an Olympus E-PM2 camera affecting both juvenile and adult NARWs (NARWC was used, while modified DJI Inspire 1 Pro quad- 2018), we projected both reproductive classes to have copters with Zenmuse X5 cameras were used in Aus- a lower body condition compared to SRWs. tralia, Argentina and New Zealand. Both UAV types were equipped with an Olympus M Zuiko 25 mm f1.8 lens to minimize picture distortion. Measurement 2. MATERIALS AND METHODS accuracies of both the APH-22 and the Inspire 1 Pro systems have been estimated at 99.9% (Durban et al. 2.1. Data collection 2015, Dawson et al. 2017, Christiansen et al. 2018). Christiansen et al. (2018) further quantified the Aerial photographs of right whales were taken measurement errors of the Olympus 25 mm lens using non-invasive (Christiansen et al. 2016b) un - when flying at different altitudes ranging from 5 to manned aerial vehicles (UAVs) in 4 locations: the 120 m and measuring a known sized object on land.

Fig. 1. Location of the 4 study regions, with sample sizes and dates of data collection provided for each location. All whales were measured on their breeding grounds, except for immature and mature North Atlantic right whales (*), which were measured on their feeding grounds (top filled circle for the North Atlantic) 4 Mar Ecol Prog Ser 640: 1–16, 2020

Fig. 2. (A) Example aerial photograph of a right whale used to measure body condition, showing the positions of the measurement sites used in the study. W: width. (B) Right whale body volume as a function of body length for the 4 sampled locations. The solid line represents the back-transformed fitted values of the linear model. (C) Log-log relationship between body volume and body length for the 4 sampled locations, with the solid line representing the fitted values of the linear model. N = 523 whales

Their results showed that within the altitude range ple measurements from the same individuals. To used in this study (17.8 and 55.1 m) the mean meas- avoid pseudo-replication, only a single measurement urement error was 0.7 cm (SD = 0.5, n = 50) with a for each whale was used, which was selected ran- maximum of 1.6 cm. Since the measurement errors domly (using a random number generator in R) from were not influenced by the altitude of the UAV, dif- the best photographs of that individual. We judged ferences in sampling altitude between locations did this to be a less biased approach compared to using not bias measurements. When photographing whales, the average measurements from repeated photo- a camera gimbal ensured that the camera of the UAV graphs of the same individuals (taken over several was always facing down at a near-perfect 90° angle. days), as the latter might introduce temporal varia- Only photographs of adequate quality, when the tion in body measurements, and also cause hetero- whale was lying flat at the surface, were used for geneity in measurement errors between individuals analyses (for details, see Christiansen et al. 2018). (animals with single or multiple measurements) and From the aerial photographs, individual whales were populations. identified based on the unique callosity patterns on Following the protocol of Christiansen et al. (2018), their heads (Payne et al. 1983). The Australian, New each photograph was graded (given a score of 1 Zealand and North Atlantic data sets included multi- [good quality], 2 [medium quality] or 3 [poor quality]) Christiansen et al.: Body condition of right whale populations 5

on several attributes, including degree of body roll, measured in this study (11.72 m). A NARW mother degree of body arch, body pitch (vertically), body with an older calf (>4 mo old), measured on the Cape length measurability and body width measurability. Cod Bay feeding ground, was removed from the Only photographs with a roll, arch or pitch <3 were analyses. Similarly, adults with body volumes similar used in the analyses. To account for variation in body to or exceeding that of lactating females with newly length and width measurability between photo- born calves (calf body length ~5 m), were removed graphs, we ran a sensitivity analysis where the body from the analyses as these likely represent late-stage length and width of each individual whale were ran- pregnant females. The absolute and relative body domly varied within the confidence interval given by width (body width/body length) of right whales at its length and width measurability scores (for details, each measurement site was compared between loca- see Christiansen et al. 2018). By repeating this pro- tions and for each reproductive class, using linear cess 1000 times, and refitting the final models in the models in R 3.5.3 (R Core Team 2019) (Figs. S2 & S3). analyses, the effect of the length and width measure- From the body length and width measurements, ment errors on the model parameters could be evalu- body volume was calculated using the methods of ated (see Fig. S1 in the Supplement at www. int-res. Christiansen et al. (2018). By assuming a circular com/articles/ suppl/m640p001_ supp .pdf). cross-sectional body shape, the body volume of the whales was estimated by dividing the body of the whale into 18 frustum segments (1 between each

2.2. Morphometric measurements and classification width measurement) and calculating the volume (Vs) of reproductive classes of each segment s using the formula of a truncated cone (Christiansen et al. 2018): The total body length of the whales (tip of the lower 1 jaw to the notch of the tail fluke) and their body VhrrRR= π 22++ (1) S 3 () widths (at 5% increments along the entire body axis of the animal), were measured (Fig. 2A), using a where the height (h) is given by the distance custom written graphical user interface (GUI) in between width measurements (5% of the body MATLAB (Dawson et al. 2017). The GUI accounted length), and the smaller (r) and larger radii (R) corre- for distortion of the camera lens through photogram- spond to half of the smaller and larger width meas- metric calibration (for details, see Dawson et al. 2017). urements, respectively (for details, see Christiansen

All photographs were measured by a single experi- et al. 2018). Total body volume (VTotal) of the whales enced researcher, thus minimizing any potential inter- was then estimated by summing the volumes of the observer bias. Image scale was established from the different frustum segments (Christiansen et al. 2018): known focal length in combination with precise alti- S tude data. For the North Atlantic data, altitude of the VVTotal = ∑ S (2) APH-22 was measured using an inbuilt GPS (mean s=1 error = 0.05 m; Durban et al. 2015) or a LightWare Similar to Christiansen et al. (2018), the body vol- SF11/C laser range finder (mean error = 0.02 m; ume of immature, mature and lactating females was Dawson et al. 2017). For the SRW populations, the calculated between 25 (the end of the head region) altitude of the Inspire 1 Pro was measured using the and 80% of their body length, which corresponds to same type of range finder. While the accuracy of the the metabolically most active region of baleen altimeters used might have differed slightly, the whales (Lockyer et al. 1985, Miller et al. 2012, Chris- width to length ratio of the whales was not affected tiansen et al. 2013, 2016a, 2018). Since the width to by this, and hence this did not bias the body condi- length ratio of calves is known to increase across tion estimates. their entire body axis during the first month of their Each whale was classified into 1 of 4 reproductive lives (Christiansen et al. 2018), the body volume of classes: calves (<4 mo of age), immature (juveniles), calves was calculated from the tip of their rostrum mature (non-lactating adults) and lactating females. down to 80% of their body length. Calves and lactating females were distinguished based on their close association with each other on the calving/breeding grounds. Immature and mature 2.3. Body condition index whales were separated based on their body length, using a threshold value of 12.0 m, which was based An animal’s body condition provides a measure of on the body length of the smallest lactating female its energy balance, health and quality (Jakob et al. 6 Mar Ecol Prog Ser 640: 1–16, 2020

1996, Peig & Green 2009, 2010). Although body con- ured individuals. If we had modelled BCI separately dition can be expressed through any physiological for each reproductive class (i.e. fitting a separate index that represents an individual’s energy reserves length-to-volume model for each reproductive class) (Hanks 1981, Millar & Hickling 1990), it generally it would have resulted in a slight shift in the inter- refers to the relative size of energy stores compared cept (mean body condition) for each reproductive with structural components (commonly the length) of class, but would not have influenced the effect of the body (Green 2001). Consequently, an individual’s location within each reproductive class (Fig. S5). body condition strongly influences its survival and Finally, although our BCI was based on a cross- reproductive success (Gaillard et al. 2000, Clutton- sectional sample of the population (a single meas- Brock & Sheldon 2010). Christiansen et al. (2018) urement representing a single whale), it correlated showed that the body volume of lactating SRW strongly with the BCI calculated from repeated females at the time they give birth will determine the measurements of the same whales (available for the amount of energy they invest in their calves (i.e. the Australian data set, see Fig. S6). rate of decline in maternal body volume), and consequently calf growth rates (i.e. rate of increase in calf body volume). Based on this, we calculated the body 2.4. Differences in body condition between condition index (BCI) of individual right whales from locations the residuals of the log-log relationship (to account for non-linear relationships) between body volume To determine if NARWs were in poorer body con- and body length, divided by the expected (or pre- dition compared to the 3 southern populations, we dicted) body volume for the individual (to standard- developed linear models in R 3.5.3. Right whale body ized BCI across body size, Christiansen et al. 2013, condition (response variable) was modelled as a 2016a, 2018): function of location (explanatory variable). Separate models were run for each reproductive class (calves, BVobs,− BV exp, BCI = ii (3) immature, mature and lactating females). i BV exp, i During the breeding season, lactating females where BVobs,i is the observed body volume of whale i have finite energy reserves to support their own 3 in m , and BVexp,i is the expected body volume of metabolic needs and the growth of their calf (Lockyer whale i in m3, given by the log-log relationship 2007, Christiansen et al. 2018). Consequently, lactat- between body volume and body length: ing females decline in body condition through the breeding season as their calves grow in size (Chris- log() BVexp, ii= α + β × log() BL (4) tiansen et al. 2016a, 2018). To account for the temporal variation in body condition of lactating females α where BLi is the body length of whale i, and and (i.e. female body condition declining with increased β re present the intercept and slope parameters, re - calf length), calf body length was included as a spectively, of the linear relationship between body covariate in the model. Similarly for calves, the effect volume and body length for all locations combined. of maternal body length and condition on calf body A positive BCI means that an individual was in condition was investigated. Other covariates in - relatively better condition than an average individ- cluded day of the year (DOY; with the North Atlantic ual of the same body length, whereas a negative data converted to austral DOY by adding 183 d) and BCI means that the individual was in relatively body length. However, collinearity (high correlation) poorer condition. To demonstrate that our BCI was between location and DOY, as well as between loca- independent from the absolute size (length) of the tion and body length, resulted in only location being individual, we calculated the body condition of all included in the final model for juvenile and adult measured whales using both the absolute (body right whales. To investigate the effect of location on length and widths, in metres) and relative body body length, separate linear models were developed morphometrics (body length and widths, in pixels). for each reproductive class. The 2 approaches yielded nearly identical BCIs Model validation included testing for homoge- 2 (F1,521 = 739902, p < 0.001, R = 0.999, Fig. S4), and neous residuals (by plotting model residuals against showed that our metric accounted for potential the fitted model values), examining normality of structural differences (i.e. body length) between residuals (from frequency histograms of residuals) animals. The BCI of each individual was calculated and influential points and outliers (by calculating from the length-to-volume relationship of all meas- leverage scores and Cook’s distance, respectively). Christiansen et al.: Body condition of right whale populations 7

2.5. Validation of cross-species comparison on Fortune et al. 2012) and SRWs (from the Argentina population) (Table 1). To enable comparison of the body condition of In regards to life history characteristics, Soviet the NARW and the SRW, their physiology and life catch data between 1951 and 1971 showed that history need to be very similar, so that the genetic female SRWs reach sexual maturity around 12.5 m variance does not exceed the environmental vari- in body length (Tormosov et al. 1998). Similarly, ance. To validate our cross-species comparison, we Sharp et al. (2019) classified NARW adults as in - therefore compared the body shape, size and dividuals >9 yr of age, which, based on recent age-to- life history characteristics of NARWs and SRWs length curves, corresponds to a body length around (Table 1). 12.5 m (Table 1). Further, we found that the minimum To compare the structural body shape of NARWs body length of lactating SRWs in this study was and SRWs, we first measured the relative body width 11.72 m, which was very similar to the minimum of the whales, and compared their head (0−25% body length of lactating NARWs, which was 11.86 m body length from the rostrum) and tail regions (Table 1). In regards to birth size, our smallest meas- (80−100% body length from the rostrum). With both ured NARW calf was 3.9 m in body length (which is these areas being mainly structural (Brodie 1975), within their predicted birth range of 4.22 ± 0.4 m; and not part of the metabolically active body area for Fortune et al. 2012), which was very similar to the right whales (between 25 and 80% of body length, smallest SRW calf at 4.1 m body length (Table 1). Christiansen et al. 2018), any genetic difference in Huang et al. (2009) also presented similar calf wean- the external body width between species should be ing lengths (8.78 m vs. 8.26 m) and female asymptotic visible in those areas. We found no difference in the lengths (17.8 m vs. 16.6 m) for NARWs and SRWs body width of the head or tail region of the whales (Table 1). (Table 1, Fig. S3). Further, although the site-specific body widths of NARWs and SRWs differed across the metabolically active body area of the whales, their 3. RESULTS overall body shapes were very similar (Fig. S3). Finally, Christiansen et al. (2019) showed that the The body volume of 523 right whales was success- relationship between body mass (or body volume) fully measured between 2016 and 2018 in the 4 study and length was very similar between NARWs (based locations (Fig. 1). There was a strong linear relation-

Table 1. Comparison of body shape, size and life history characteristics between the North Atlantic right whale (NARW) and the southern right whale (SRW). BL: body length

Structural body shape/ NARW SRW Source life history characteristic

Relative width (%BL ± SE) of head Calf = 20.0 ± 0.67 Calf = 20.4 ± 0.28 This study (Fig. S3) (20% BL from rostrum) Immature = 20.1 ± 0.21 Immature = 20.8 ± 0.18 Mature = 19.5 ± 0.15 Mature = 20.8 ± 0.22 Lactating = 19.8 ± 0.45 Lactating = 20.9 ± 0.13 Relative width (%BL ± SE) of tail Calf = 5.5 ± 0.15 Calf = 5.0 ± 0.16 This study (Fig. S3) (80% BL from rostrum) Immature = 4.2 ± 0.16 Immature = 4.0 ± 0.08 Mature = 4.3 ± 0.10 Mature = 3.8 ± 0.13 Lactating = 3.6 ± 0.19 Lactating = 3.9 ± 0.10 Female length (m) at sexual maturity 12.5 12.5 Tormosov et al. (1998), Sharp et al. (2019) Minimum length (m) of lactating females 11.7 11.9 This study Female asymptotic length (m) 17.8 16.6 Huang et al. (2009) Minimum length (m) at birth 3.9 4.1 This study Length at weaning (m) 8.8 8.3 Huang et al. (2009) Weight (kg) at birth (BL = 4 m) 940 870 Christiansen et al. (2019), Fortune et al. (2012) Weight (kg) at weaning (BL = 8.5 m) 7,830 7,970 Christiansen et al. (2019), Fortune et al. (2012) Weight (kg) at sexual maturity (BL = 12.0 m) 20,680 21,940 Christiansen et al. (2019), Fortune et al. (2012) 8 Mar Ecol Prog Ser 640: 1–16, 2020

ship between body volume (BV) and body length 17.6% units lower compared to lactating SRW

(BL) on the log-log scale (F1,521 = 28953, p < 0.001, females in Australia (mean = 11.2%, SE = 2.1), New R2 = 0.982, Fig. 2C): Zealand (mean = 15.6%, SE = 3.4) and Argentina (mean = 8.3%, SE = 6.5), respectively (Figs. 3 & 4D). log() BVexp, ii= − 4.38+× 3.01 log() BL (5) The body length analyses showed that lactating NARW females (mean = 13.2 m, SE = 0.29) were on

Lactating females in the North Atlantic were in average 96, 54 and 48 cm shorter (F3,157 = 5.07, p = poorer body condition compared to the southern pop- 0.002, R2 = 0.088) than Australian (mean = 14.2 m, 2 ulations (F3,156 = 5.11, p = 0.002, R = 0.072, Fig. 3; and SE = 0.32), Argentinian (mean = 13.8 m, SE = 0.30) see Fig. S7 and model 6 in Table S1). The body con- and New Zealand females (mean = 13.7 m, SE = dition of lactating females from all 4 populations 0.36), respectively (Fig. 4C). Consequently, the mean decreased as the calf grew in size (i.e. body length) absolute body volume (mean = 27.9 m3, SE = 2.90) of through the breeding season (F1,156 = 42.02, p < 0.001, lactating NARWs was significantly (F3,157 = 6.10, p < R2 = 0.197), at a rate of 6.62% BCI m−1 calf length 0.001, R2 = 0.104) lower than for SRWs in Australia (Fig. 4D, model 6 in Table S1). If we account for the (mean = 39.9 m3, SE = 3.11), New Zealand (mean = body length of their calves (i.e. fix calf length to 6.0 m 38.1 m3, SE = 3.56) and Argentina (mean = 36.1 m3, in the model), the body condition of lactating NARW SE = 2.99), at a magnitude of 11.99, 10.22 and females (mean = −9.4%, SE = 4.8) was 20.5, 24.9 and 8.22 m3, respectively.

Fig. 3. (A) Predicted body condition values from the best fitting models for right whale calves (model 1 in Table S2 in the supplement), immature whales, mature whales and lactating females (model 6 in Table S1), as a function of location. (B) Predicted body condition values for right whales from Argentina, Australia, New Zealand and the North Atlantic, as a function of reproductive class. Error bars represent 95% confidence intervals. All whales were measured on their calving/breeding grounds, except for immature and mature North Atlantic right whales, which were measured on their feeding grounds. For lactating females, the full model also included calf body length as an explanatory variable, with maternal body condition declining significantly with calf body length (Fig. 4C). In the partial effect plot shown here, calf length was fixed at 6 m, which represents the mean body length of calves measured in this study. Sample sizes for all reproductive classes are given in Fig. 1 Christiansen et al.: Body condition of right whale populations 9

Fig. 4. (A) Partial effect plot of right whale calf length as a function of maternal body length, with maternal body condition fixed at 0. (B) Partial effect plot of right whale calf length as a function of maternal body condition, with maternal body length fixed at 14.0 m (the mean length of lactating females in the data set). The solid lines represent the fitted values of the best fitting linear model (Model 6 in Table S3) and the dashed lines represent 95% confidence intervals. (C) Maternal body length between locations. (D) Right whale maternal body condition as a function of calf body length for different locations. The solid lines represent the fitted values of the best fitting linear model (model 6 in Table S1). N = 161 lactating females with calves

NARW calves did not show signs of being in poorer Table S3). The full model explained 23.3% of the condition compared to SRW calves, and the body variance in the data. length of calves (a rough proxy for time since birth) There was a difference in body condition of mature did not vary significantly between locations. Further, right whales (males and non-lactating females) be - 2 the day of sampling did not vary significantly between tween locations (F3,90 = 25.06, p < 0.001, R = 0.455, NARWs (after correcting the time of year with 183 d Fig. 3; Figs. S7 & S8), with North Atlantic adults between the Northern and Southern hemisphere) (mean = −16.7%, SE = 2.0) being 27.9, 18.9 and 8.9% and SRWs in Australia and New Zealand. Instead, we units lower in condition compared to individuals from found that Australian calves (mean = −5.7%, SE = Argentina (mean = 11.2%, SE = 3.3), New Zealand 2 2.2) were significantly (F3,157 = 9.93, p < 0.001, R = (mean = 2.2%, SE = 4.0) and Australia (mean = 0.159, Fig. 3; Fig. S7 and Table S2) thinner than New −7.8%, SE = 3.6), respectively (Fig. 3). In addition to Zealand (mean = 13.5%, SE = 3.7) and Argentinian being in poorer condition, the average body length of calves (mean = 3.7%, SE = 1.2). mature NARW (mean = 12.9 m, SE = 0.13) was lower 2 Body length of calves was positively related to the (F3,90 = 6.07, p < 0.001, R = 0.168) compared to 2 length of their mothers (F1,158 = 12.5, p < 0.001, R = Argentina (mean = 13.9 m, SE = 0.22), New Zealand 0.061) at a rate of 0.302 m (SE = 0.082) per m increase (mean = 13.4 m, SE = 0.26) and Australia (mean = in maternal length (Fig. 4A; model 6 in Table S3). In 13.3 m, SE = 0.24) (Fig. S9B). addition, maternal body condition was negatively We found that immature NARWs (mean = −13.1%, correlated (slope parameter = −2.82 m, SE = 0.474) SE = 2.9) were in significantly poorer condition (F3,103 = 2 2 with calf body length (F1,158 = 35.5, p < 0.001, R = 4.30, p = 0.007, R = 0.111, Fig. 3; Fig. S7) than juve- 0.172), since maternal body condition decreased as niles in New Zealand (mean = −1.2%, SE = 3.4), the calf grew in body length (Fig. 4B; model 6 in Australia (mean = −2.4%, SE = 4.2) and Argentina 10 Mar Ecol Prog Ser 640: 1–16, 2020

(mean = −2.9%, SE = 3.4). On average, the BCI of tigate the relationship between maternal body condi- immature NARWs was 11.9, 10.7 and 10.2% units tion and calf length (Fig. 4D). We found that the lower than juveniles from New Zealand, Australia and absolute maternal cost of producing a similar sized Argentina, respectively (Fig. 3). Juvenile NARWs calf (the slope parameter) was similar across popula- (mean = 11.2 m, SE = 0.19) were on average longer tions, while the absolute maternal body condition at a 2 (F3,103 = 7.03, p < 0.001, R = 0.170) than juveniles in given calf length (the intercept parameter) was sig- Argentina (mean = 10.5 m, SE = 0.22) and New Zealand nificantly lower for NARW females. Assuming that (mean = 10.5 m, SE = 0.22), and similar in size to Aus- NARW calves were growing at a slower rate com- tralian juveniles (mean = 11.3, SE = 0.27) (Fig. S9A). pared to SRW calves, the observed difference in The results from our sensitivity analysis showed maternal condition could be due to a difference in that all body condition model parameter values were the age of calves, with NARW calves being relatively robust to measurement errors resulting from differ- older at a given body length compared to SRW ences in picture quality (body length and width calves. With NARW females having less energy measurability) (Fig. S1). reserves available to invest in their calf, this could result in them having to wean their calf at a smaller size. While weaning size is positively correlated to 4. DISCUSSION pup survival in pinnipeds (McMahon et al. 2000), this relationship is unknown in baleen whales. Alterna- Like most baleen whales, right whales rely heavily tively, NARW females might compensate for their on stored energy for reproduction, particularly dur- lower rate of offspring investment by extending the ing lactation (Lockyer 1981, Miller et al. 2012, Chris- lactating period longer into the succeeding feeding tiansen et al. 2018). While the body condition of season, when they are able to supplement their own NARWs has declined during the last 3 decades (Rol- body condition (and hence also their offspring invest- land et al. 2016), this study provides the first compar- ment) by concurrent feeding. While this strategy ison with healthy (i.e. growing) SRW populations. In would likely result in a longer inter-calving interval agreement with our main hypothesis, we found that for NARW females, since they would need more time NARW juveniles, adults and lactating females were to replenish their energy stores, it would not lead to a all in significantly poorer body condition compared to reduction in calf survival. the SRW populations. Our results were robust to Lactating NARW females were also shorter in body measurement errors resulting from variation in pic- length than the 3 southern populations. This was not ture quality (body length and width measurability). the result of morphological differences (different The largest difference in body condition was for asymptotic body lengths) between the 2 species, lactating females, with NARW females being on since whaling and stranding records show no species average 21% units lower than the 3 SRW popula- difference in body length (Tormosov et al. 1998, tions. To put this into perspective, the body condition Moore et al. 2004, Huang et al. 2009, Fortune et al. of lactating females decreased by about 19% units 2012). With the absolute body volume of right whales during the first 3 mo of lactation, assuming a calf being largely determined by their body length growth rate of 3.2 cm d−1 (Christiansen et al. 2018). (Christiansen et al. 2018), lactating NARW females This early lactation period is considered the most likely have less energy available to invest in their energetically costly part of the reproductive cycle in calves, which again will negatively affect calf growth baleen whales, since females are still relying on rates. If we use Christiansen et al.’s (2018) relation- stored energy reserves during this time, while their ship between maternal investment (rate of body calf is growing rapidly in size (Lockyer 1981, Miller et volume loss) and maternal body length and condition al. 2012, Christiansen et al. 2016a, 2018). In support for SRW in Australia, the magnitude of difference for of this, we found that the body condition of lactating lactating NARW (20.5% unit lower body condition females was generally better than that of juveniles and 96 cm shorter body length) equates to a loss in and adults (Fig. 3). A compromised body condition maternal rate of investment of 50% (rate of decline in during this critical time period means that NARW maternal body volume: North Atlantic = 0.063 m3 d−1; females have considerably less energy available to Australia: 0.126 m3 d−1). Determining the lower invest in their calves, which is known to negatively threshold in body condition at which lactating fe - influence calf growth rates (Christiansen et al. 2018). males will no longer be able to energetically support While we did not have data to directly investigate their calves should be the aim of future research, as calf growth rates for NARWs, we were able to inves- well as identifying the threshold below which fertility Christiansen et al.: Body condition of right whale populations 11

(the probability of a female becoming pregnant) and Argentina (11°C). In contrast, juveniles, adults and pregnancy (the ability to complete gestation) is sig- lactating females all rely on their own energy re- nificantly compromised. serves during the breeding season, and so their sur- Despite the smaller size and poorer body condition vival and reproductive success is likely to be more of their mothers, NARW calves did not show signs of closely linked to their body condition, whereas heat being in poorer condition compared to SRW calves. loss is likely to be less important due to their overall This is contrary to our hypothesis, which were based larger body size (lower surface area to volume ratio) on the findings of Christiansen et al. (2016a), who and thicker blubber layer (more insulation). reported a positive relationship between calf body Similar to lactating females, mature NARWs were condition and maternal body condition in humpback in poorer body condition and smaller in size (i.e. body whales Megaptera novaeangliae. While this lack of length) than the 3 SRW populations. Miller et al. effect could be due to a difference in the timing of (2011, 2012) found a similar difference in blubber sampling between locations, this is unlikely since the thickness and body width of NARWs during the sum- body length of calves (a rough proxy for time since mer feeding season and SRWs in South Africa during birth) did not vary significantly between locations. It the winter breeding season. The observed difference hence seems that a reduction in maternal body con- could be due to variations in the timing of sampling, dition in right whales does not lead to a reduction in with NARW adults being sampled early in their sum- calf body condition, although it could still be sup- mer feeding season (when their energy reserves are pressing calf growth in length. still low from the previous breeding season) while We found that Australian calves had a significantly SRW adults were measured during their winter lower body condition compared to the other SRW breeding season (when they still have much of their populations. This was unexpected, since lactating energy reserves remaining). However, the magni- females in Australia had similar body condition as tude of the difference in body condition between females in New Zealand and Argentina (Fig. 3), and NARW and SRW adults in our study (Argentina = so we anticipated their calves to have similar condi- 27.9% units, Australia = 8.9% units, New Zealand = tion. However, the lower body condition of Aus- 18.9% units) was similar or exceeded the observed tralian calves does not seem to be correlated with variation in body condition within locations (95% lower calf survival, since the population is growing at confidence range: Argentina = 15.2%, Australia = a similar rate (5.55%, Bannister 2016) to the New 9.3%, New Zealand = 13.6%, North Atlantic = 5.6%), Zealand population (5−7%, Carroll et al. 2013), and suggesting that variation in the time of sampling also has similar inter-calving intervals (Australia: alone cannot explain the observed difference in body 3.33 yr, Burnell 2001; New Zealand: 3.31 yr, David- condition of adults between locations (Fig. 3). Fur- son et al. 2017). Further, only 4.2% of all measured ther, had the NARW adults been measured towards calving intervals in Australia were 2 yr (Charlton the end of the feeding season (when they are at their 2017), an indication that females lost their calf early peak body condition), the fact that they are still in in lactation (Marón et al. 2015), compared to 8.9% in poorer BCI compared to SRW adults is even more New Zealand (Davidson et al. 2017). It thus seems alarming. Repeated sampling of NARW and SRW that the body condition of calves, within the range of adults on their feeding grounds, to determine the values observed in this study, is not linked to their rate of fattening, is needed to accurately quantify the survival. Logically, calves should starve to death if magnitude of the difference in BCI between adults their body condition falls below a critical limit where from the 2 species. they can no longer afford to maintain homeostasis. From the measured mature NARWs of known sex, However, assuming that their mothers can support 36.1% (13 of 36) were females, which, based on their them with sufficient energy, in the form of milk, to reproductive cycle, should have been either in a preg- support their basic metabolic needs, baleen whale nant or resting state (non-pregnant, non-lactating). A calves do not necessarily need to build up large fat reduction of body reserves in pregnant females can reserves to survive. Instead, the lower body condition result in less energy available for the foetus, which in (higher surface area to volume ratio) of Australian minke whales Balaenoptera acutorostrata has been calves might be an adaptation to the relatively shown to result in smaller (i.e. shorter) foetuses warmer waters (lower heat loss) experienced on their (Christiansen et al. 2014). However, given that no breeding grounds (15°C, sea surface temperature NARW calves were born in the 2017−2018 breeding on 1 August, www.meteoblue.com) compared to season (NARWC 2018), we can conclude that the New Zealand (6.1−7.7°C, Rayment et al. 2015) and measured adult females in this study were either 12 Mar Ecol Prog Ser 640: 1–16, 2020

resting, had a failed pregnancy or lost their calf in body condition can still be significant (Réale et al. shortly after giving birth (before they could be 1999, Merilä et al. 2001). Further, differences in sighted). In resting females, lower body condition salinity and prey depth between locations (feeding suggests that females are taking longer to recover grounds) could influence the optimal body shape (fat from reproduction and nursing than right whales in to muscle ratio) to achieve neutral buoyancy during other populations. This could help explain the sub- foraging (Narazaki et al. 2018). However, our find- stantially longer calving interval of NARW females ings show that the structural body shape of NARWs (>7 yr) versus the SRW populations (~3.3 yr) (Burnell and SRWs is very similar, while published records 2001, Cooke et al. 2003, Davidson et al. 2017, Pettis et demonstrate similar body sizes and life history char- al. 2020). By combining aerial photogrammetry (to acteristics of the 2 species (Tormosov et al. 1998, determine BCI) and breath sampling (to determine Huang et al. 2009, Sharp et al. 2019). Further, our reproductive status), future research should aim to data show that the body condition (and hence vol- assess whether a reduction in the BCI of NARW fe- ume) needed to produce a similar-sized offspring males is negatively affecting their fertility (their abil- was the same for lactating females across popula- ity to become pregnant), pregnancy (their ability to tions (Fig. 4D), suggesting that the body energy con- complete gestation), offspring survival (their ability to tent was similar across species and locations. Finally, energetically support their calf) and/or the time of re- differences in water temperature between locations covery (their ability to deposit energy) from calving. (both on the breeding and feeding grounds) could As for the adult whales, the lower body condition of influence the optimal body shape (and hence BCI) for immature NARWs could be partly due to variations in minimizing heat loss. While this could explain the the timing of sampling between the North Atlantic observed difference in body condition between right and the southern populations. Younger NARW juve- whale calves, heat loss is unlikely to lead to a popu- niles (1−4 yr) have also been found to have lower lation difference in body condition of juveniles and body condition (i.e. blubber thickness) compared to adult whales, due to their significantly larger body older juveniles (5−8 yr) (Miller et al. 2011). Potential size (lower surface area to volume ratio and thicker age differences between locations are unlikely to blubber layer), and ability to tolerate a wide variation explain the lower body condition of juvenile NARWs in temperatures across their spatial range (between in this study, which on average were longer than subtropical and subpolar zones). juveniles in Argentina and New Zealand, and similar The observed differences in body condition be - in size to Australian juveniles. Although the implica- tween the NARW and the SRW populations are most tion for vital rates is hard to determine, poorer body likely to result from differences in the exposure to condition in juvenile NARWs could reduce the energy anthropogenic factors. While the 3 SRW populations available for growth. This, in turn, could delay sexual examined reside in relatively remote and unim- maturation, which in baleen whales is strongly influ- pacted environments, the home range of the NARW enced by body size (Sigurjónsson et al. 1990). All else overlaps with heavily developed coastal areas, the being equal, delayed sexual maturation would act greatest lobster and crab trap and line densities and to slow the population growth rate. A comparison of some of the world’s busiest shipping lanes (Moore length-at-age growth curves between locations would 2019). Despite management actions, ship strikes re - help determine if NARWs are growing at a slower main responsible for ongoing right whale mortalities rate compared to the southern populations. This high- in the North Atlantic (Moore et al. 2004, Knowlton et lights the value of long-term monitoring projects with al. 2012, van der Hoop et al. 2013, Sharp et al. 2019); well-studied photo-identified individuals, for which however, morbidity and mortality due to entangle- age can be accurately determined. ment has become the predominant source of diag- With the NARWs being genetically isolated from nosed trauma to NARWs since 2010 (NOAA 2018). the SRWs (Rosenbaum et al. 2000, Gaines et al. 2005), This increasing entanglement in fishing gear is a it is possible that some of the observed difference in major threat to NARWs; more than 83% of individu- BCI between the 2 species derives from genetic als carry scars from at least 1 entanglement, and divergence and local adaptations to different envi- 15.5% of the population is entangled every year ronmental conditions (e.g. different water tempera- (Knowlton et al. 2012). The additional drag, buoy- ture and prey availability). Although traits closely ancy and impeded foraging ability caused by various associated with fitness, such as body condition, gen- fishing gear leads to significant increases in the erally show low heritabilities (Mousseau & Roff 1987, energy expenditure of right whales (Cassoff et al. Kruuk et al. 2000), the genetic component of variance 2011, van der Hoop et al. 2016, 2017). The cumula- Christiansen et al.: Body condition of right whale populations 13

tive energetic costs and stress resulting from repeti- West Coast Aboriginal Corporation for access to Aboriginal tive and prolonged interactions with fishing gear can lands, and C. Charlton and F. Vivier for help in the field; this paper represents HIMB and SOEST contribution numbers lead to substantial reductions in body condition (Rol- 1788 and 10919; New Zealand: Crew and expedition mem- land et al. 2012, Schick et al. 2013, Pettis et al. 2017, bers on board ‘Polaris II’; Argentina: J. Graham, M. Riccia- van der Hoop et al. 2017), which could result in rdi, A. Fernández Ajó, M. Di Martino, C. Adrian Diaz and R. reproductive failure and even death (Moore et al. Soley for help in the field, the Ferro family for access to their property and Instituto de Conservación de Ballenas for 2004, Robbins et al. 2015, Rolland et al. 2016). Anthro- logistical support. Finally, we thank P. Palsbøll and 3 anony- pogenic noise (e.g. from shipping) increases stress mous reviewers for their constructive comments which in NARWs, which carries energetic costs (Rolland et helped to improve this manuscript. al. 2012). Finally, climate-associated changes in right whale LITERATURE CITED prey (the copepod Calanus finmarchicus) availability Aguilar A (1986) A review of old Basque whaling and its and distribution in the North Atlantic are believed to effect on the right whales (Eubalaena glacialis) of the reduce the rate of energy intake, body condition and North Atlantic. Rep Int Whaling Comm 10: 191−199 consequent calving rates (Miller et al. 2011, Meyer- Albon SD, Mitchell B, Staines BW (1983) Fertility and body weight in female red deer: a density-dependent relation- Gutbrod et al. 2015, Meyer-Gutbrod & Greene 2018). ship. J Anim Ecol 52: 969−980 The smaller absolute body size (i.e. body length) of Arnbom T, Fedak MA, Boyd IL (1997) Factors affecting NARW adults and lactating females further suggests maternal expenditure in Southern elephant seals during that the high mortality rate of females in the North lactation. Ecology 78: 471−483 Atlantic might have skewed their age range to a sig- Atkinson SN, Ramsay MA (1995) The effects of prolonged fasting of the body composition and reproductive success nificantly lower average age than in SRWs (Pace et of female polar bears (Ursus maritimus). Funct Ecol 9: al. 2017). This would deprive NARW females of the 559−567 advantage of larger size as shown by Christiansen et Bannister JL (2016) Monitoring population dynamics of al. (2018). The 2017−2018 NARW breeding season ‘western’ right whales off southern Australia 2015−2018. Report to the Marine Biodiversity Hub, National Envi- highlighted the severity of the situation, when not a ronmental Science Programme, University of Tasmania, single calf was born into the population (NARWC Hobart 2018). Unless their situation improves soon, the Bannister JL, Pastene LA, Burnell SR (1999) First record of ongoing decline of NARWs will result in them movement of a southern right whale (Eubalaena australis) between warm water breeding grounds and the becoming another of the growing list of cetaceans Antarctic Ocean, south of 60°S. Mar Mamm Sci 15: (including vaquita Phocoena sinus, Pennisi 2017; 1337−1342 Maui dolphin Cephalorhynchus hectori maui, Pala Best PB (1994) Seasonality of reproduction and the length of 2017; Gulf of Mexico Bryde’s whale Balaenoptera gestation in southern right whales Eubalaena australis. J edeni, Corkeron & Kraus 2018) at serious risk of Zool (Lond) 232: 175−189 Boltnev AI, York AE (2001) Maternal investment in northern extinction. fur seals (Callorhinus ursinus): interrelationships among mothers’ age, size, parturition date, offspring size and Acknowledgements. Research permits: North Atlantic: NOAA sex ratios. J Zool (Lond) 254: 219−228 17355-01; Australia: DEWNR M26501-2 & MR00082-3-V; Bowen WD, Iverson SJ, Boness DJ, Oftedal OT (2001) Forag- New Zealand: Department of Conservation 50094-MAR; ing effort, food intake and lactation performance depend Argentina: Sub Secretaría de Conservación y Áreas Protegi- on maternal mass in a small phocid seal. Funct Ecol 15: das 43-SsCyAP/18 and Dirección de Fauna y Flora Silvestres 325−334 106/2018.SsG-M.P. Animal ethics permits: North Atlantic: Brodie PF (1975) Cetacean energetics, an overview of intra- Woods Hole Oceanographic Institution Institutional Animal specific size variation. Ecology 56: 152−161 Care and Use Committee; Australia and Argentina: Mur- Burnell SR (2001) Aspects of the reproductive biology doch University O2819/16. UAV permits: North Atlantic: FAA and behavioral ecology of right whales off Australia. MOU 2016-ESA-3-NOAA; Australia: 1-YC6NP-03. Funding: J Cetacean Res Manag 2: 89−102 North Atlantic: NOAA NA14OAR4320158; Australia: US Burnell SR (2008) Estimates of demographic parameters of Office of Naval Research Marine Mammals Program (Award southern right whales off Australia. Rep SC/60/BRG12. No. N00014-17-1-3018), the World Wildlife Fund for Nature International Whaling Commission, Cambridge Australia and a Murdoch University School of Veterinary Carroll EL, Childerhouse SJ, Fewster RM, Patenaude NJ and Life Sciences Small Grant Award; New Zealand: New and others (2013) Accounting for female reproductive Zealand Antarctic Research institute (NZARI 2016-1-4), cycles in a superpopulation capture-recapture frame- Otago University and NZ Whale and Dolphin Trust; Argen - work. Ecol Appl 23:1677−1690 tina: National Geographic Society (Grant number: NGS- Cassoff RM, Moore KM, McLellan WA, Barco SG, Rotstein 379R-18). Thanks to: North Atlantic: NOAA’s Office of Mar- DS, Moore MJ (2011) Lethal entanglement in baleen ine and Aircraft Operations and Aircraft Operations Center, whales. Dis Aquat Org 96: 175−185 D. LeRoi and W. Perryman for technical support; Australia: Charlton CM (2017) Population demographics of southern Aboriginal Lands Trust, Yalata Land Management and Far right whales (Eubalaena australis) in Southern Australia. 14 Mar Ecol Prog Ser 640: 1–16, 2020

PhD thesis, Curtin University, Bentley Fortune SME, Trites AW, Perryman WL, Moore MJ, Pettis Christiansen F, Vikingsson GA, Rasmussen MH, Lusseau D HM, Lynn MS (2012) Growth and rapid early develop- (2013) Minke whales maximise energy storage on their ment of North Atlantic right whales (Eubalaena feeding grounds. J Exp Biol 216: 427−436 glacialis). J Mammal 93: 1342−1354 Christiansen F, Víkingsson GA, Rasmussen MH, Lusseau D Gaillard JM, Festa-Bianchet M, Delorme D, Jorgenson J (2014) Female body condition affects foetal growth in a (2000) Body mass and individual fitness in female ungu- capital breeding mysticete. Funct Ecol 28: 579−588 lates: bigger is not always better. Proc R Soc B 267: Christiansen F, Dujon AM, Sprogis KR, Arnould JPY, Bejder 471−477 L (2016a) Noninvasive unmanned aerial vehicle provides Gaines CA, Hare MP, Beck SE, Rosenbaum HC (2005) estimates of the energetic cost of reproduction in hump- Nuclear markers confirm taxonomic status and relation- back whales. Ecosphere 7: e01468 ships among highly endangered and closely related right Christiansen F, Rojano-Doñate L, Madsen PT, Bejder L whale species. Proc R Soc B 272:533−542 (2016b) Noise levels of multi-rotor unmanned aerial Green AJ (2001) Mass/length residuals: measures of body vehicles with implications for potential underwater condition or generators of spurious results? Ecology 82: impacts on marine mammals. Front Mar Sci 3: 277 1473−1483 Christiansen F, Vivier F, Charlton C, Ward R, Amerson A, Hanks J (1981) Characterization of population condition. In: Burnell S, Bejder L (2018) Maternal body size and condi- Fowler CW, Smith T (eds) Dynamics of large mammal tion determine calf growth rates in southern right populations. John Wiley, New York, NY, p 47−73 whales. Mar Ecol Prog Ser 592:267−281 Huang SL, Chou LS, Ni IH (2009) Comparable length at Christiansen F, Sironi M, Moore MJ, Di Martino M and oth- weaning in cetaceans. Mar Mamm Sci 25: 875−887 ers (2019) Estimating body mass of free-living whales International Whaling Commission (2013) Report of the using aerial photogrammetry and 3D volumetrics. Meth- workshop on southern right whales. Doc SC/65A/Rep05. ods Ecol Evol 10: 2034−2044 International Whaling Commission, Cambridge Clutton-Brock T, Sheldon BC (2010) Individuals and popula- Jackson JA, Carroll EL, Smith TD, Zerbini AN, Patenaude tions: the role of long-term, individual-based studies of NJ, Baker CS (2016) An integrated approach to historical animals in ecology and evolutionary biology. Trends Ecol population assessment of the great whales: case of the Evol 25: 562−573 New Zealand southern right whale. R Soc Open Sci 3: Cooke J, Rowntree V, Payne R (2003) Analysis of inter-annual 150669 variation in reproductive success of South Atlantic right Jakob EM, Marshall SD, Uetz GW (1996) Estimating fitness: whales (Eubalaena australis) from photo-identifications a comparison of body condition indices. Oikos 77: 61−67 of calving females observed off Península Valdés, Knowlton AR, Hamilton PK, Marx MK, Pettis HM, Kraus SD Argentina, during 1971-2000. Rep SC/55/O23. Interna- (2012) Monitoring North Atlantic right whale Eubalaena tional Whaling Commission, Cambridge glacialis entanglement rates: a 30 yr retrospective. Mar Cooke J, Rowntree V, Sironi M (2015) Southwest Atlantic Ecol Prog Ser 466: 293−302 right whales: interim updated population assessment Kraus SD, Hamilton PK, Kenney RD, Knowlton AR, Slay CK from photo-ID collected at Peninsula Valdez, Argentina. (2001) Reproductive parameters of the North Atlantic right Rep SC/66a/BRG/23. International Whaling Commis- whale. J Cetacean Res Manag (Spec Issue) 2: 231−236 sion, Cambridge Kraus SD, Kenney RD, Mayo CA, McLellan WA, Moore MJ, Corkeron P, Kraus SD (2018) Baleen whale species on brink Nowacek DP (2016) Recent scientific publications cast of extinction for first time in 300 years. Nature 554: 169 doubt on North Atlantic right whale future. Front Mar Sci Crespo EA, Pedraza SN, Dans SL, Svendsen GM, Degrati M, 3: 137 Coscarella MA (2019) The southwestern Atlantic south- Kruuk LEB, Clutton-Brock TH, Slate J, Pemberton JM, ern right whale, Eubalaena australis, population is grow- Brotherstone S, Guinness FE (2000) Heritability of fitness ing but at a decelerated rate. Mar Mamm Sci 35: 93−107 in a wild mammal population. Proc Natl Acad Sci USA Davidson AR, Rayment W, Dawson SM, Webster T, Slooten 97: 698−703 E (2017) Estimated calving interval for the New Zealand Lockyer C (1981) Estimation of the energy costs of growth, southern right whale (Eubalaena australis). N Z J Mar maintenance and reproduction in the female minke Freshw Res 52:372−382 whale, (Balaenoptera acutorostrata), from the Southern Dawbin WH (1986) Right whales caught in waters around hemisphere. Doc SC/32/Mi19. The Scientific Committee south-eastern Australia and New Zealand during the of the International Whaling Commission, Cambridge nineteenth and early twentieth centuries. Rep Int Whal- Lockyer C (1987) The relationship between body fat, food ing Comm 10: 261−268 resource and reproductive energy costs in North Atlantic Dawson SM, Bowman MH, Leunissen E, Sirguey P (2017) fin whales (Balaenoptera physalus). Symp Zool Soc Lond Inexpensive aerial photogrammetry for studies of whales 57: 343−361 and large marine animals. Front Mar Sci 4: 366 Lockyer C (2007) All creatures great and smaller: a study in Douhard F, Gaillard JM, Pellerin M, Jacob L, Lemaître JF cetacean life history energetics. J Mar Biol Assoc UK 87: (2017) The cost of growing large: costs of post-weaning 1035−1045 growth on body mass senescence in a wild mammal. Lockyer CH, McConnell LC, Waters TD (1985) Body condi- Oikos 126: 1329−1338 tion in terms of anatomical and biochemical assessment Durban JW, Fearnbach H, Barrett-Lennard LG, Perryman of body fat in North Atlantic fin and sei whales. Can J WL, Leroi DJ (2015) Photogrammetry of killer whales Zool 63: 2328−2338 using a small hexacopter launched at sea. J Unmanned Loudon ASI, McNeilly AS, Milne JA (1983) Nutrition and Veh Syst 3: 131–135 lactational control of fertility in red deer. Nature 302: Festa-Bianchet M (1998) Condition-dependent reproductive 145−147 success in bighorn ewes. Ecol Lett 1:91−94 Marón CF, Rowntree VJ, Sironi M, Uhart M, Payne RS, Christiansen et al.: Body condition of right whale populations 15

Adler FR, Seger J (2015) Estimating population conse- body condition from mass/length data: the scaled mass quences of increased calf mortality in the southern right index as an alternative method. Oikos 118: 1883−1891 whales off Argentina. Rep SC/66a/BRG/1. International Peig J, Green AJ (2010) The paradigm of body condition: a Whaling Commission, Cambridge critical reappraisal of current methods based on mass McMahon CR, Burton HR, Bester MN (2000) Weaning mass and length. Funct Ecol 24:1323−1332 and the future survival of juvenile southern elephant Pennisi E (2017) After failed rescue effort, rare porpoise in seals, Mirounga leonina, at Macquarie Island. Antarct extreme peril. Science 358: 851 Sci 12:149−153 Pettis HM, Rolland RM, Hamilton PK, Knowlton AR, Burgess Merilä J, Kruuk LEB, Sheldon BC (2001) Natural selection EA, Kraus SD (2017) Body condition changes arising on the genetical component of variance in body condi- from natural factors and fishing gear entanglements in tion in a wild bird population. J Evol Biol 14: 918−929 North Atlantic right whales Eubalaena glacialis. Endang Meyer-Gutbrod EL, Greene CH (2018) Uncertain recovery Species Res 32:237−249 of the North Atlantic right whale in a changing ocean. Pettis HM, Pace RM, Hamilton PK (2018) North Atlantic Glob Chang Biol 24: 455−464 Right Whale Consortium 2018 Annual Report Card. Meyer-Gutbrod EL, Greene CH, Sullivan PJ, Pershing AJ North Atlantic Right Whale Consortium, Boston, MA. (2015) Climate-associated changes in prey availability www.narwc. org/uploads/1/1/6/6/116623219/2018report drive reproductive dynamics of the North Atlantic right _cardfinal.pdf whale population. Mar Ecol Prog Ser 535:243−258 Pettis HM, Pace RM, Hamilton PK (2020) North Atlantic Millar JS, Hickling GJ (1990) Fasting endurance and the Right Whale Consortium 2019 Annual Report Card. evolution of mammalian body size. Funct Ecol 4: 5−12 North Atlantic Right Whale Consortium, Boston, MA. Miller CA, Reeb D, Best PB, Knowlton AR, Brown MW, www.narwc.org/uploads/1/1/6/6/116623219/2019 report Moore MJ (2011) Blubber thickness in right whales final.pdf Eubalaena glacialis and Eubalaena australis related with R Core Team (2019) R: a language and environment for sta- reproduction, life history status and prey abundance. tistical computing. R Foundation for Statistical Comput- Mar Ecol Prog Ser 438: 267−283 ing, Vienna Miller CA, Best PB, Perryman WL, Baumgartner MF, Moore Rayment WJ, Dawson S, Webster T (2015) Breeding status MJ (2012) Body shape changes associated with repro- affects fine-scale habitat selection of southern right whales ductive status, nutritive condition and growth in right on their wintering grounds. J Biogeogr 42: 463−474 whales Eubalaena glacialis and E. australis. Mar Ecol Réale D, Festa-Bianchet M, Jorgenson JT (1999) Heritability Prog Ser 459: 135−156 of body mass varies with age and season in wild bighorn Moore MJ (2019) How we can all stop killing whales: a pro- sheep. Heredity 83: 526−532 posal to avoid whale entanglement in fishing gear. ICES Reeves R, Breiwick J, Mitchell E (1999) History of whaling J Mar Sci 76:781−786 and estimated kill of right whales, Balaena glacialis, in Moore MJ, Knowlton AR, Kraus SD, McLellan WA, Bonde the northeastern United States, 1620−1924. Mar Fish Rev RK (2004) Morphometry, gross morphology and available 61: 1−36 histopathology in North Atlantic right whale (Eubalaena Reeves RR, Rolland RM, Clapham PJ (2001) Causes of repro- glacialis) mortalities (1970-2002). J Cetacean Res Manag ductive failure in North Atlantic right whales: new 6: 199−214 avenues of research. Report of a workshop held 26-28 Mousseau TA, Roff DA (1987) Natural selection and heri- April 2000, Falmouth, Massachusetts. Ref Doc 01-16. tability of fitness components. Heredity 59:181−197 Northeast Fisheries Science Center, Woods Hole, MA Narazaki T, Isojunno S, Nowacek DP, Swift R and others Robbins J, Knowlton AR, Landry S (2015) Apparent survival (2018) Body density of humpback whales (Megaptera of North Atlantic right whales after entanglement in fish- novaengliae [sic]) in feeding aggregations estimated ing gear. Biol Conserv 191: 421−427 from hydrodynamic gliding performance. PLOS ONE 13: Rolland RM, Parks SE, Hunt KE, Castellote M and others e0200287 (2012) Evidence that ship noise increases stress in right NOAA (2018) North Atlantic right whale (Eubalaena whales. Proc R Soc B 279: 2363−2368 glacialis) stock assessment report. Natl Ocean Atmos Rolland RM, Schick RS, Pettis HM, Knowlton AR, Hamilton Adm: 13−29. www.fisheries.noaa.gov/webdam/ download/ PK, Clark JS, Kraus SD (2016) Health of North Atlantic 82311292 right whales Eubalaena glacialis over three decades: NARWC (North Atlantic Right Whale Consortium) (2018) from individual health to demographic and population Identification Database 09/04/2018. North Atlantic Right health trends. Mar Ecol Prog Ser 542: 265−282 Whale Consortium, Anderson Cabot Cent Ocean Life Rosenbaum HC, Brownell RL Jr, Brown MW, Schaeff C and New Engl Aquarium, Boston, MA others (2000) World wide genetic differentiation of Pace RM III, Corkeron PJ, Kraus SD (2017) State-space Eubalaena: questioning the number of right whale spe- mark-recapture estimates reveal a recent decline in cies. Mol Ecol 9:1793−1802 abundance of North Atlantic right whales. Ecol Evol 7: Schick RS, Kraus SD, Rolland RM, Knowlton AR and others 8730−8741 (2013) Using hierarchical Bayes to understand move- Pala C (2017) New Zealand’s endemic dolphins are hanging ment, health, and survival in the endangered North by a thread. Science 355: 559 Atlantic right Whale. PLOS ONE 8: e64166 Payne R, Brazier O, Dorsey EM, Perkins JS, Rowntree VJ, Sharp SM, McLellan WA, Rotstein DS, Costidis AM and oth- Titus A (1983) External features in southern right whales ers (2019) Gross and histopathologic diagnoses from (Eubalaena australis) and their use in identifying individ- North Atlantic right whale Eubalaena glacialis mortali- uals. In: Payne R (ed) Communication and behavior of ties between 2003 and 2018. Dis Aquat Org 135: 1−31 whales. Westview Press, Boulder, CO, p 371−445 Sigurjónsson J, Halldórsson SD, Konráðsson A (1990) New Peig J, Green AJ (2009) New perspectives for estimating information on age and reproduction in minke whales 16 Mar Ecol Prog Ser 640: 1–16, 2020

(Balaenoptera acutorostrata) in Icelandic waters. Doc Smith J, Moore MJ (2016) Drag from fishing gear entan- SC/42/NHMi27. The Scientific Committee of the Inter- gling North Atlantic right whales. Mar Mamm Sci 32: national Whaling Commission, Noordwijkerhout 619−642 Skogland T (1984) The effects of food and maternal condi- van der Hoop J, Corkeron P, Moore M (2017) Entanglement tions on fetal growth and size in wild reindeer. Rangifer is a costly life-history stage in large whales. Ecol Evol 7: 4: 39−46 92−106 Stephens PA, Boyd IL, McNamara JM, Houston AI (2009) Wheatley KE, Bradshaw CJA, Davis LS, Harcourt RG, Hin- Capital breeding and income breeding: their meaning, dell MA (2006) Influence of maternal mass and condition measurement, and worth. Ecology 90: 2057−2067 on energy transfer in Weddell seals. J Anim Ecol 75: Tormosov DD, Mikhaliev YA, Best PB, Zemsky VA, 724−733 Sekiguchi K, Brownell RL Jr (1998) Soviet catches of Williams R, Vikingsson GA, Gislason A, Lockyer C, New L, southern right whales Eubalaena australis, 1951−1971. Thomas L, Hammond PS (2013) Evidence for density- Biological data and conservation implications. Biol Con- dependent changes in body condition and pregnancy serv 86:185−197 rate of North Atlantic fin whales over four decades of van der Hoop JM, Moore MJ, Barco SG, Cole TVN and oth- varying environmental conditions. ICES J Mar Sci 70: ers (2013) Assessment of management to mitigate anthro- 1273−1280 pogenic effects on large whales. Conserv Biol 27: 121−133 Yablokov AV (1994) Validity of whaling data. Nature 367: van der Hoop JM, Corkeron P, Kenney J, Landry S, Morin D, 108

Appendix. Complete list of authors’ addresses

Fredrik Christiansen1,2,3,*, Stephen M. Dawson4, John W. Durban5, Holly Fearnbach6, Carolyn A. Miller7, Lars Bejder2,3,8, Marcela Uhart9,10, Mariano Sironi9,11,12, Peter Corkeron13, William Rayment4, Eva Leunissen4, Eashani Haria3, Rhianne Ward14, Hunter A. Warick3, Iain Kerr15, Morgan S. Lynn5, Heather M. Pettis13, Michael J. Moore16

1Aarhus Institute of Advanced Studies, Høegh-Guldbergs Gade 6B, 8000 Aarhus C, Denmark 2Zoophysiology, Department of Biology, Aarhus University, C.F. Møllers Alle 3, 8000 Aarhus C, Denmark 3Centre for Sustainable Aquatic Ecosystems, Harry Butler Institute, Murdoch University, Murdoch, 6150 Western Australia, Australia 4Department of Marine Science, University of Otago, Dunedin 9054, New Zealand 5Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA, La Jolla, CA 92037, USA 6SeaLife Response, Rehabilitation and Research, Seattle, WA 98126, USA 7Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02543 USA 8Marine Mammal Research Program, Hawaii Institute of Marine Biology, University of Hawaii at Manoa, HI 96744, USA 9Southern Right Whale Health Monitoring Program, Puerto Madryn, Chubut 9120, Argentina 10School of Veterinary Medicine, University of California Davis, Davis, CA 95616, USA 11Instituto de Conservación de Ballenas, Buenos Aires 1429, Argentina 12Diversidad Biológica IV, Universidad Nacional de Córdoba, Córdoba 5000, Argentina 13Anderson Cabot Center for Ocean Life, New England Aquarium, Boston, MA 02110, USA 14Centre for Marine Science and Technology, Curtin University, Bentley, 6102 Western Australia, Australia 15Ocean Alliance, Gloucester, MA 01930, USA 16Biology Department, Woods Hole Oceanographic Institution, Woods Hole, MA 02543, USA

Editorial responsibility: Per Palsbøll, Submitted: June 17, 2019; Accepted: March 16, 2020 Groningen, The Netherlands Proofs received from author(s): April 17, 2020